Search

Multiple disasters induced by hurricanes such as floods, strong winds, and heavy rainfall have always been a significant threat to coastal infrastructures. However, few studies have attempted to model the dependence between hazards and study its impact on multi-hazard risks—these cognitive deficiencies inevitably affect the control of risks. This paper addresses these difficulties by first using the copulas to capture the dependence between hurricane-induced floods and strong winds based on synthetic multi-hazard events. Moreover, another two types of dependence, regression-based and mutually independent, are constructed to compare the impact of different dependence behaviors on the facility’s failure. Second, the frequency of concurrent events and a scenario-based fragility function are convolved to quantify the multi-hazard risk of storage tanks. Subsequently, the annualized failure risk (AFR) and the mean return period of failure (MRPF) are used as two metrics for the multi-hazard risk. The case study results demonstrate that the BB1 copula can accurately describe the change in dependence across the full range of the multivariate distribution, avoiding overestimation or underestimation of equipment failure risks. The application of risk zoning provides decision-makers with more intuitive and convenient tools for planning and layout, highlighting the practicability of the proposed multi-hazard risk assessment method.

In this study, the diatomite (DE) was firstly dispersed by the electron beam bombardment (EB) to form the modified diatomite (MDE), using as the carrier of the Fe doped carbon nitride (FGCN). The MDE/FGCN demonstrates broader light adsorption in the visible region and low recombination of photogenerated electron-hole, according to the results of spectroscopy techniques. Reduction of hexavalent chromium (Cr(VI)) in the aqueous solution was used to illustrate the photocatalytic ability of MDE/FGCN. Results showed that the MDE/FGCN-1:3 (Fe, 10%) composites with EB dose of 30 kGy exhibited stronger reduction efficiency with 98.3% of Cr(VI) in 100 min, which was almost 16.92 times higher than that of pure GCN for the reaction rate. Additionally, lower pH values could promote the Cr(VI) reduction, and the MDE/FGCN displayed great stability after four runs. Lastly, the electron and heat effects of EB were creatively used to analyze the dispersion mechanism of MDE. The higher photocatalytic ability of MDE/FGCN could be attributed not only the dispersion of MDE but also the charge trapping by Fe. This study will broaden the application of physical irradiation in the modification of carriers for the photocatalyst, which can be applied in the treatment of environmental pollutants.

A strong maximum ebb current (>3 m/s) in the upper water column was observed at Station CS3 in the middle of the North Passage of the Yangtze River Estuary during the wet season, which was higher than either its upstream or downstream counterparts. To better understand the mechanisms and factors causing the strong ebb current, a vertical one-dimensional (1-D) model was used to conduct a diagnostic study. The model used time series of observed tidal amplitudes, vertical salinity, and suspended sediment concentration (SSC) profiles to compute the density and turbulence. Two tunable parameters, the tidal amplitude attenuation coefficient (i.e., the phase lag) and the background surface pressure gradient that represents the net pressure gradient induced by the freshwater discharge and baroclinic effect, were used to determine the best match with the observed high velocity amplitudes in addition to the bottom roughness height. Three hypotheses of possible causes are tested: (1) the large freshwater discharge, (2) the bottom stratification effects (which were caused by a possible high near-bed suspended sediment gradient), and (3) the unique location of the CS3 station that was influenced by local geometry. The findings show that neither of the first two factors has much influence on the pronounced ebb velocities. Instead, the energy loss caused by the change of channel geometry and a maximum convex bathymetry in the North Passage of the Yangtze River Estuary are the main reasons behind the extremely high observed ebb current velocity profiles. The high near-bottom SSC and gradient located within 0.5 m above the bed only slightly alters the velocity profiles. This 1-D model is convenient for testing a different hypothesis and for coupling with other selected variables to account for the floc size distributions in future studies.

Polymer supports play an important role in noble metal loaded catalysts. In this study, palladium clusters stabilized by several polyamides have been widely studied in the oxidation of benzyl alcohol under airflow. The different kinds of polyamides (PADs) were synthesized with aromatic diamines and trimesoyl chloride by mechanical grinding method and solvent-heat method. Polyamide support was synthesized through mechanical grinding method. And the laminar structure and abundant amide bonds of polyamide can help obtain ultrafine Pd clusters with a mean particle size of 1.76 nm. The Pd0.5@PAD-2-M effectively catalyzed benzyl alcohol into benzaldehyde with a conversion of 96.1 % and turnover number (TON) of 1441 under 90 °C in water which was 2.5 times higher than the commercial Pd/C catalyst. High TON value of 6505 was also achieved in solvent-free system. This work reveals the advantages of laminar structure of polyamide in dispersing Pd clusters and promoting benzyl alcohol conversion. Besides, the carrier properties of amphipathy and substrate adsorption were also investigated to illustrate the significance of interfacial effect for laminar catalyst in heterogeneous catalysis.

The Multi-Azimuth UV Imager (MAVI) is designed and built by Chinese scientific institute independently. The MAVI is comprised of the ultraviolet ring imager and the ultraviolet forward imager. One track of image of ultraviolet ring imager contains hundreds of standard image frames arranged along the track direction and possesses relatively higher data volume. To produce the standard level-one product rapidly, the entire procedure of geometric correction was decomposed into parameter parsing, data reading into memory, model of geometric correction, and data exporting to hard disk. Among these four steps of ultraviolet ring imager image processing, the construction and calculating of geometric correction model is the most time consuming. There exist several for loops in steps of geometric positioning and vector calculating. Simultaneously, the majority of C code of geometric positioning and vector calculating should be rewritten to adapt the CUDA GPU in this practice. Experimental results show that parallelized geometric correction could reach a speedup of 3.2 through image partition along the column direction on GPU. A better speedup result of URI image processing would be achieved through parallelizing all for loops in every step.

Deep learning technology such as convolutional neural networks (CNN) has achieved outstanding results in the field of crops classification for remote sensing images. The way of land cover or crop types remote sensing classification using CNN is mainly pixel-based classification which is often affected by the phenomenon of “salt and pepper”. In order to reduce this effect, an object-oriented crops classification method based on CNN is proposed in this paper. By combining image segmentation technology and CNN model, we use this method to obtain the results of crops classification from Sentinel-2A multi-spectral remote sensing images in Yuanyang County, Henan Province, China. The experiment show that, compared with the pixel level classification based on CNN which only consider the spectral and temporal characteristics of the crops, the method we proposed comprehensively utilizes more detailed information such as spectral feature, texture feature, spatial relationship, and color space. Thus, it gains a better discriminability for some specific crop and achieves higher classification accuracy.

The main contribution of this article is to solve the problem of detection of larger scale differences. Aiming at the problem of small target detection, we can better use the underlying features of the convolution network to construct the hyper convolution feature to achieve better detection and recognition effect. For larger scale target, by dilated convolution operation, the context information of different scales can be integrated into high-level feature information according to different receptive fields. In this experiment, we introduce the lightweight convolutional network, SqueezeNet, as the basic feature network. The network has small size, fast training speed and strong expression ability. In the experiment environment of single Titan X GPU card, the distribution of the migrated dataset can be better studied by increasing the size of batch images during training. After the pre-training of the VOC dataset, the migration training was carried out in the remote sensing image dataset, and the mAP of the detection of the 12 targets reached 0.937205, which reached a better level of detection result.

In this paper, we design and develop two SLAM (simultaneous localization and mapping) methods: FTF-VO (frame-to-frame visual odometry) and MF-VO (multi-frame visual odometry) for robotic localization based on stereo cameras. In both VO methods, we proposed an ARFM-based 3D reconstruction method and calculated the position of the camera using PNP, after that BA (bundle adjustment) can be used to optimize the final camera position. We tested FTF-VO and MF-VO using KITTI dataset, with BA and without BA respectively. The result demonstrates that MF-VO with BA has the highest precision so that it is suitable for the scenarios requiring high precision. Compared with FTF-VO with BA, MF-VO achieves better localization precision while requires less running time, so FTF-VO with BA is recommended for medium localization precision. FTF-VO is the best choice for low precision.

To improve the recognition and detection accuracy and recall of objects in unnatural images, we use remote sensing video image data to detect low-quality and small targets. Based on this, we propose the use of a deconvolution network and hyper features to control convolutional feature quality. We call this approach the Quality Deconvolution Single Shot Detector (QDSSD) detection model. Through the frame-by-frame annotation of the video data from Jilin Satellite No. 1, we propose a CSU-RSI-Video dataset, with no fewer than 30 targets per image frame. We published the data for researchers to do experiments. To achieve small target detection, we enrich the information by gradually adding the underlying detail features to the upper layers and deconvolve the high-layer information to obtain stable detailed features for target detection. The empirical results show that, among the detected problems of low-quality small targets, the improved QDSSD network has better detection capability, and the detection effect is the best for many small targets that are close to each other. In the CSU-RSI-Video dataset, the QDSSD model’s mAP achieves 0.90227 for a single target. Comparing to the You Only Look Once (YOLO) model, the result is still superior in accuracy.

Dealing with natural disasters, such as debris flow, typically involves teleoperation of an unmanned robot. Three-dimensional (3D) map information of an environment is especially useful for teleoperating the unmanned robot system to perform unmanned construction and auto-investigation. Recently, structure from motion (SfM) using an unmanned aerial vehicle (UAV) equipped with monocular cameras has been proposed as a general method to build 3D map information. However, absolute slope and scale information for the built 3D map cannot be obtained when applying general SfM technology without the deployment of ground control points (GCPs). Therefore, we propose a method that can recover absolute slope and scale information for a 3D map. To this end, we apply sensor fusion technology that combines inertial measurement unit (IMU) data from an unmanned ground vehicle (UGV) and image data from the camera mounted on a UAV. Experimental results show that our estimation scheme is able to successfully represent the slope and scale information along with the 3D map.

The wide band images acquired from the Tiangong-2 space laboratory covers many spectral bands such as visible light, shortwave infrared and thermal infrared. These high-quality images can be used for space science experiments such as earth observation. In this paper, we use CNN (convolutional neural networks) to extract the spectral features of different landcover from the wide band images, then identify the early rice and the late rice accurately in Huarong County, Hunan Province, China. With advanced techniques such as deep learning, the spatial distribution information of crops can be effectively obtained from the wide band images which can provide data services for agricultural production management.

At the South Branch of the Changjiang Estuary near Baimao Shoal, two in situ approaches were used to estimate the settling velocity, ws , of suspended, fine-grained sediments. The first approach wa...

Laboratory experiments were conducted to compare the performances of four different instruments for measuring suspended cohesive sediment concentrations (SSCs). Among these instruments, two were different models of optical backscatter sensor (i.e., OBS3+ and OBS5+), one was an acoustic Doppler velocimeter (MicroADV), and the last was a laser infrared optical sensor developed at Hohai University, China (HHU-LIOS). Sediments collected from the Yangtze River Estuary and a commercially available kaolinite were selected to check the responses of these four instruments. They were placed in an aqueous solution, and the SSCs were changed within a range from about 10 mg/L to 30 g/L to demonstrate the effective measuring ranges for each instrument. For the kaolinite suspension, the results showed that the HHU-LIOS has a much larger linear response range (0.1 to 13 g/L) when compared with other devices (the MicroADV: 0.1 to 1.35 g/L; the OBS3+: 0.1 to 5 g/L; and the OBS5+: 0.1 to 1 g/L). For the sediment sample from the Yangtze River Estuary, the results showed that the HHU-LIOS has a much wider usable range (from 0.1 to 25 g/L), compared with those for the others. It also showed that a combination of using a HHU-LIOS for the quadratic increasing range up to 25 g/L and the OBS5+ for the clear decreasing output are able to measure the SSCs of the estuarine mud up to 50 g/L.